Background: Myocardial perfusion imaging (MPI) is a useful method for risk assessment in patients with diabetes mellitus (DM) and chronic kidney disease (CKD), but these patients have a residual risk that is not accounted for by MPI. The objective of this study is to determine whether the heart rate response (HRR) to adenosine has an incremental prognostic value to MPI in high-risk patients. Methods: The study group included 879 (age 61 ± 13 years, 48% women, 58% white, 40% DM, 49% CKD) consecutive patients who underwent adenosine MPI. Chronic kidney disease was defined as an estimated glomerular filtration rate <60 mL/min per 1.73 m2 or dialysis replacement therapy. An HRR <10% (change from baseline) was considered blunted. The outcome of interest was overall mortality. Results: During a follow-up period of 40 ± 14 months, 212 patients (24%) died. Patients with DM (23.4% ± 16.3% vs 29.4% ± 21.4%, P <.0001) and CKD (22.7% ± 17.6% vs 30.5% ± 20.4%, P <.0001) had lower HRR as compared with patients without DM and CKD, respectively. A blunted HRR was associated with increased mortality in the overall population and in those with DM and CKD and helped in risk stratification when added to traditional MPI findings. In a Cox regression model, a blunted HRR was the strongest predictor of mortality (hazard ratio 2.8, P <.0001) and provided additional prognostic data to MPI (hazard ratio 1.9, P <.0001) after controlling for age, gender, race, history of myocardial infarction, DM, CKD, β-blocker use, and presence of chest pain. Conclusions: A blunted HRR to adenosine is an independent predictor of poor outcome, adds incremental value to MPI, and helps in better risk stratification in high-risk patient groups. © 2011 Mosby, Inc. All rights reserved.